Impact of Occupancy Previewing Mismatch on Nonlinear Model Predictive Control for Indoor Air Quality Management

Occupancy in an office can significantly influence Indoor Air Quality parameters. A reliable control of Indoor Air Quality represents a crucial requirement in the Heating, Ventilation and Air Conditioning field. For this purpose, specific modelization and control techniques can be used in order to guarantee the desired level of the involved parameters. A nonlinear Model Predictive Control strategy is used in the present paper in order to manage carbon dioxide level in an office through natural ventilation. This controller is located on an upper level with respect to other controllers, e.g., temperature ones. Occupancy and wind speed are taken into account in the control problem formulation as measured input disturbance variables. The impact of occupancy previewing mismatch on carbon dioxide control is assessed through tailored simulations, which considered exact knowledge, underestimation and overestimation of the occupancy.